The main purpose of this SIG is to provide a European, interdisciplinary forum for all people who are interested in research and development of intelligent information agents (I2A). This concerns, in particular, researchers from the following areas:
- Database and Information Systems
- Information Retrieval
- Artificial Intelligence
- CSCW (Computer Supported Collaborative Work)
- HCI (Human Computer Interaction), HAI (Human Agent Interaction)
There is a strong interest in applying results from research on intelligent information agents also in industry, e.g. for improving commercial (multi-)database products and/or tools for electronic retail commerce in the internet. Research on intelligent information agents is an inherent interdisciplinary issue and entails topics such as: Systems and Architectures of information agents. Application of Advanced Database and Knowledge-Base Technology for the development of intelligent information agents. Construction and use of Ontologies for multi-agent information gathering in distributed, hetereogeneous information sources. Communication among information agents on the Internet. Access to Data Warehouses, and Integration of Information from heterogeneous sources by autonomous (collaborating) information agents. Learning, interaction and organisation of information agent systems for information discovery in the Internet/Web. Security aspects for information servers and agents in the Internet. Human-Agent Interaction for information agents in distributed WWW applications. Visualization of User Information Spaces discovered by Information Agents. Issues of programming (systems of) information agents for the Internet. Electronic Commerce and Rational Information Agents in the Internet. Advanced protocols for negotiation and electronic commerce useful for information agents in electronic markets (virtual agent market places) on the Internet. Game-theoretic and other microeconomic principles for rational information agents. The I2A-SIG promotes synergetic efforts on applying software agent technology to database and information systems in open environments, such as the Internet or large corporate intranets. Specifically, it is expected to establish new cooperations among research groups in the related areas mentioned above, but also to strengthen existing contacts and focus scattered efforts for research on and development of intelligent information agents. This includes promotion of collaborative projects and cross fertilisation of ideas between nodes of the AgentLink NoE which are active in R&D of information agents.
Some other long-term goals are
- Understanding the benefits, constraints and limitations of applying agent technology in the (multi-)database and information systems area, and vice versa;
- Investigating the impact of R&D on information agents on users' everyday lives in the Infosphere (like today's relatively simple ShopBots or SearchBots). This entails discovering the impact of, among other things:
- Rational information agents on the Internet/Web on the real-world economy and market through their commercial use by information providers, and their acceptance by customers.
- Adaptive information agents on, for example, Human-Computer Interaction R&D, and vice versa. In particular, this means the use of techniques for effective, cost-/time-efficient, multimedia representation of and navigation in individually available information spaces in order to provide users with fast content-based information searches, and to enable users to interact with their personal (or other) information agents in a convenient way.
- Discovering methodologies for software development and implementation of different kinds of (systems of) information agents;
- Developing benchmarks for measurement of performance and efficiency of (systems of) information agents in all kinds of modern information environments;
- Realizing the development and (prototypical) implementation of novel types of (systems of) information agents in the Internet or Web as a result of joint efforts among participants of the SIG.
|
